Age Correlations and an Old Earth: Part II.

Is it a good idea to raz someone with a close friend in the admin ranks?

By "species dependent" do you mean that certain species do not exhibit the phenomena I inquired about?

It's way beyond what I know. I think only some species are inclinded toward false rings. I think (only think again) that most are somewhat susceptable to the missing rings though. If the conditions are bad enough.

I just have found new dating service. Spam removed by AdminJarAre there any new features?This message has been edited by AdminJar, 12-10-2004 09:05 AM

heh ... always gotta test the limits eh?how was the skiing? I never got to whistler, but skied mount washington on vancouver island ... powder snow down the backside bowl, moguls through the forests ... fun stuffs. (I was younger then).(you're friends with your alter ego? do you talk often? is this something we should be concerned about?)

Hi RAZD,In regards to this article...What is the procedure used to calibrate the C14 method by the varve method? I know they carbon date (determine the C14/C12 ratio, right?) a sample fossil from a varve. I know they determine the age of the sample fossil's varve by counting. So, the fossil has a varve age and a carbon age, of course. But do you (or someone else) happen to know the *procedure* by which the carbon age is calibrated by the varve age? (Does the article say, and I'm just missing it?) So, every varve age should be given in a range (e.g., 12,000 years B.P. +/- 2000 years), right? Which means the C14 age, which the varve age is calibrating, cannot be any more precise than to give an age within a 4000 year range (+/- 2000 years), right? In this study, are the initial C14 ages of the fossils in the varves known? I understand that the present ratio between C14/C12 (or whatever) can be measured, but how does one determine the *initial* C14 age of a given sample?Are they suggesting that their original 4000 year range ( +/- 2000 years) for any given varve might be a 10,000 year range (+/- 5000 years) in some cases?Also, a little Googling on diatoms makes me think that to assume a once-a-year diatom layer is a bit simplistic. Diatom blooms, apparently, are affected by nitrogen, ammonium, phosphate (all common plant-fertilizer ingredients, btw) and silicate levels in the water as well as by the amount of sunlight they receive. (Check out this article: Florida Bay Watch Report: the Plume and the Bloom)Did these guys do a separate, detailed study on the present conditions that could affect diatom activity in this lake? Or, did they, perhaps, refer to such a study of this lake? Considering the variables involved in diatom activity (and I never was able to find information on possible common sources of massive diatom destruction {a statement that could be misleading removed by TheLiteralist}), how is it reasonable to assume these varves form annually in the present? Considering the variables involved in diatom activity, how is it reasonable to assume that the varves have been forming annually every year for the past 20,000 or 40,000 years?This has been quite challenging, BTW. I haven't even gotten to the other things...I'm still stuck on lake varves. Thanks.This message has been edited by TheLiteralist, 12-11-2004 06:26 AM

I don't know if this adds anything to the discussion or not, but, according to an online Smithsonian Magazine, diatoms live about 6 days. (Seems like a rather short life span to me).Even if it doesn't enter into the discussion, it's cool to know.

Literalist: welcome back to the land of the living.I'll combine both posts in the answers here, and take them a little out of order to answer the more salient points first. What you have are alternating layers of diatoms and very fine clay, and this is what makes the layers annual.During the summer months the diatoms bloom and live and die in great numbers, their little shells dropping to the bottom and layering it in a rather unsteady but constant rain. Even in a slow growth summer (day\week) period, the production of diatoms far exceeds the sedimentation rate of the fine clay, so diatoms are the predominant sediment during the summer months.During the winter months the diatoms shut down and live a much slower life, so production of shell material slows dramatically, and the predominant sediment is clay settling out of the water from the runoff into the lake. Clay does not settle out fast, so a layer of clay means that there is no significant diatom activity: it is a winter layer.What is being counted is winters, if you will, rather than years, so yes, one thing that could throw an odd year in or out is a "nuclear winter" type event (meteor) or the like - statistically irrelevant. This is done by setting the C14 date equal to the varve date. This seems a little simplistic, but that is all that is involved: both dates have to come from the same time, we have a very high assurance of the date from the varves, the error from counting varve layers is less than the errors in the C14 (because we don't know the initial N14 levels)The thing for you is not to get hung up on the calibration issue (that is for those who have no problem with C14 dating as an accurate method and are looking to increase that accuracy by this calibration), but to look at the correlation between the two methods: if C14 was not accurate the data points should be in a random cloud on the graph rather than a linear (with some diffusion) pattern. If you mentally remove he vertical bars and look at just the data points you see a general trend across and up the graph.This graph shows the correlation used to calibrate the C14 dates so that a C14 date can be entered on the right side, carried across to a curve drawn through the data points and then taken down to a "varve date" to get a corrected C14 age of a sample not from Lake Suigetsu: You do this by working the C14 dating calculation in reverse with the age from the known source (varves in this case) to derive what the C14_o had to have been at that time. Again it is a calibration issue rather than a correlation issue as it assumes the real date correlation is 1:1 rather than looking to see what the correlation is (and perhaps to verify if in fact there is a correlation). A perfect 1:1 correlation would be the straight line shown on the graph, and this would be true if the N14 levels in the atmosphere had been constant (and at current levels). Think of it more as a percentage rate than a constant value: +/-(2,000/40,000) is +/-5%, so the amount for the 12,000 year age would be +/-600 years. um ... , no. I believe you are mixing calibration with correlation again. The variation in calibration is not relevant to the variation in correlation. This is mixing the distance of the calibration curve from a straight line up with the variation in the data points about the calibration curve (if I understand you here).Hope this helps.Enjoy.

---

we are limited in our ability to understand
by our ability to understand RebelAAmerican.Zen[Deist
{{{Buddha walks off laughing with joy}}}

RAZD covered your questions pretty well I think but there does seem to be some rather fundamental misunderstandings in your questions so I'll try to word some answers differently to see if it helps.

quote:

---

and the accumulated error at 40,000 cal yr B.P. would be less than 2000 years, assuming no break in the sediment

---

You misread this by missing the word "accumulated". If you count 10 varves (or anything) you probably have no error at all. If you count 100 you may have an error or two, if you count 1,000 there will be more errors. They are saying that by the time they've counted all 40,000 the error is less than 2000 years. Just a note that you need to clearly separate the calibration with the correlation as noted at the end of RASD's post.This paper is almost completely concerned with calibration for those who accept that C14 dating is generally a sensible to do. The other message that you might need to look at is that the C14 dates without calibration match up to the varve counts reasonably well.(I'm assuming that your 4000 year range ( +/- 2000 is meant to be 40,000 btw)You missed in reading the paper that the larger error (2000 years at the 40,000 year mark) is for those measurements over 20,000 years because they only used one core for over 20,000 years.You missed the comment about C14 dates being 5,000 years too young for dates over 31,000 years BP. It is not suggested that the 5,000 years applies before that. They point out that they have good correlation with other measurements before that. They suggest more work is needed after that. This is, again, all part of the calibration issue.

Ned added some good points.another point to make is the relation of the early varve data to the tree ring data.there is a rather brief mention that the first 10,000 years are correlated with the tree ring data from the (I believe it is the oak) dendrochronology series.((again this paper is for people familiar with dating methods and who do not have an controversy with them, so they do not explain them in as much detail as they might for a more eclectic audience.))this is the important point: correlations. tree rings correlate to lake varves correlate to C₁₄ dating. they come from different methods and different places around the earth, and yet they have the same patterns of years and seasons (mild summers, volcanic incidents, etc, show up in the data in the different systems).this makes it increasingly hard for any system to be wrong as any explanation of errors in one system has to also show how the same pattern can emerge in another system.more on this later, when you get to the devil's cave data ...enjoy.

---

we are limited in our ability to understand
by our ability to understand RebelAAmerican.Zen[Deist
{{{Buddha walks off laughing with joy}}}

Hi RAZD,Sorry, I'm still having trouble understanding this calibration thing. For the moment, I am not concerned with the validity of the calibration method, but presently I am trying to understand HOW the calibration works in some detail (I emphasize: don't worry about whether or not I see these timing methods as valid).This graph, in particular, is incomprehensible to me: What do these data points represent exactly, and how were the data points on this graph obtained? {This question added via edit by TheLiteralist}
Also, just like in graph B of that Lake Suigetsu article, this one has different starting points and scales for each axis. Why?And then about this graph: Since the varves and tree-rings are assumed to be annual, shouldn't they {typo removed by TheLiteralist} automatically line up with the actual time line (the nice, straight, diagonal line) And since the C14 ages are set equal to the varve and tree-ring ages, shouldn't the C14 ages also be equal to the nice, straight, diagonal time line? I don't understand the how the data could deviate at all from the time line, and if they do, it would be a problem in the varves or tree-rings because you say the C14 ages are simply set equal to the varve and tree-ring ages.What am I missing? I feel like I am overlooking something obvious here.I think the answer to these questions will help my understanding of the process of calibration tremendously. On the other hand, it might not help me at all. But I think it's a good starting place, nonetheless.Thanks.This message has been edited by TheLiteralist, 12-13-2004 04:51 AMThis message has been edited by TheLiteralist, 12-13-2004 04:56 AM

Please see Message 15 from loudmouth for answers.You do ask good, careful questions.His answers are:

quote:

---

The solid dots represent the ages calculated from the C14 dates in Lake Suigetsu (Japan) and the open circles represent data points from Lake Gosciaz (Poland). This graph shows that seasonal/weather pattern differences do not affect the dating of lake varves by C14 dating, being that the two sites are on opposite sides of the world.

---

The first graph shows too different sets of C14 dates matched up to actual counts of years from two different places. They are in considerable agreement with each other.and for the second he says:

quote:

---

This graph shows the correlation between lake varves (Japan, closed circles) and coral dating through uranium/thorium dating (open circles). This is especially important because the two dating methods differ in method (varves/coral, C14/uranium) and are also in different places. Again, it is the corroboration between dating methods that solidifies the reliability of these different methods.

---

That is this is another comparison of two different methods showing considerable agreement between them. It is part of checking that the calibration work they are doing is meaningful.This message has been edited by NosyNed, 12-13-2004 11:03 AM

Literalist - a couple of points that I'd like to add to what's been said:
If you count anything at all - presidential votes in Florida, varves, tree rings, whatever - there will be some error in your count. The error will be smallish for small total numbers, or for usually yes/no things like votes, but get larger - by the nature of the counting process itself - for numbers in the thousands where the items being counted can occasionally be "hazy" or ambiguous.You might also note that the paper mentions about 50 samples from deeper in the core that dated as "infinite." This just means that so much of their 14C has decayed - like in 60,000 years or more - that the amount remaining is statistically the same as zero. The authors used this oldest stuff as their "background" - the zero 14C level that the measuring method gave if you really had 0.000000000% carbon 14 in your sample, and were just measuring the minute traces you picked up from, maybe, the rubber gloves that you handled the samples with when you picked them out of the core.

Another small note:It is only the primary literature that we can use as the "real" information. However, you are experiencing some of the problems with using it.This is intended for an audience that is very familiar with the field. It is focussed on the specifics of the study undertaken. My guess is that they would be in phone, chat and email contact with almost all of those that they would expect to read the paper and answer any number of detailed questions form them.There is a mountain of understanding and training lying just under this little pebble. It isn't surprising that it can be hard to follow if you're not used to it.The two graphs you asked about are a very small example of the kind of thing one would look for in a well written paper on a well done study. They are separate things done to check that the results are reasonable. They are part of what an expert might look for before they would pass this kind of paper in review. It is part of the process of science to do that kind of thing again and again. To come at the results over and over to see what might be wrong.As an example I attended a talk by the team leader who produced the first Bose Einstein condensate and got a Nobel for it. It seems they worked for some years to reach a successful experiment. Then, with others working on beating them too it (and the prize), they waited most of a year before publishing. It is like getting to the finish line of a marathon half and hour ahead of the rest of the racers then stopping before stepping over it. They spent that time trying to figure out what might be wrong with their results and to check as much as they could think of. Then they published.That is science. That is not what is so-called "creation science".

Good questions. Mostly about the calibration issue rather than the correlation issue, but ... Each of those points represents an organic piece found in the varves that was dated by both the C₁₄ method and by the varve layers. The varve age is measured on the horizontal axis of the graph, and the C₁₄ age is measured on the vertical axis. The different starting points just means that the varve ages from 0 to ~8700 BP and the C₁₄ ages from 0 to ~7700 BP were omitted to make the graph better fit into the published page and be able to show the critical information to best advantage (again, this is for people looking at calibration not correlation, and your primary issue is the correlation) -- this is a fairly common publishing convention. Ah, you begin to see. If the propotion of C₁₄ in the atmosphere were a constant value down through the ages, as is assumed in uncalibrated C₁₄ dating, then there would be a straight line correlation (or rather straighter ... there would still be some variation in the data ... just to keep scientists honest ),BUT because of the variation in atmospheric C₁₄ the samples had different (C₁₄)_o at the beginning of their stretch at the bottom of the pool, and in terms of doing the age calculations this is like some getting headstarts and others getting handicaps in the age game, and it jiggers their ages back and forth depending on what the actual (C₁₄)_o was way back when leaf first met water. You can think of this as a graph of the error induced by (C₁₄)_o variations, and that is why it is useful for calibrating the C₁₄ dating methods: that (C₁₄)_o is world-wide (it is atmospheric).The formula for calculating the age from C₁₄ is: where:
ln ... is the natural logarithm function
c_f ... is the percentage of C₁₄ in the sample now
c_o ... is the percentage of C₁₄ in the sample originally ⁽¹⁾
t_1/2 ... is the half-life of C₁₄ (5730 years).⁽¹⁾ - this is the (C₁₄)_o above, and if we don't know what the value was way back then, then we have to assume a value, as in that it was the same as today's value. This is where the calibration comes in ... because we do know what the real age is from the varves we can work the formula backwards to find what the real (C₁₄)_o was at the time the sample first got it's feet wet.(ps - see How Carbon-14 Dating Works | HowStuffWorks for a user friendly article on the whole methodology)Now back to that curve with the straight line: let us assume that we have another lake varve system similar to Lake Suigetsu somewhere else in the world, and we take the actual age from the varve date, figure what the (C₁₄)_o would be from the Lake Suigetsu data and use that to calculate the C₁₄ age of the samples, then ... if the calibration of the C₁₄ method holds up, I would expect the results from this second lake with corrected C₁₄ ages to look like a straight line (certainly at the scale of these graphs to the naked eye, but I would not expect it to really be a straight line). BUT if there really was no calibration of the C₁₄ method, then I would expect the data to be all over the place rather than linear. This would be a good test of the calibration of the C₁₄ method. (This would also be a test of the correlation of C₁₄ age to varve age, but that is a separate issue). It is simple, but it is not necessarily easy to see what is simple the first time you meet it (how many physicists never thought of e=mc²?).Think of it as two different ways of measuring the same thing, but there is a variation in the relationship between the two methods. Say you have an investment that grows on average 10% per year: you can track the value in american dollars and you can track the value in canadian dollars -- you will get different "performance" data from the two systems due to the variation in value of the canadian dollar compared to the value of the american dollar. If you make a graph with value in american dollars along the x-axis and value in canadian dollars along the y-axis you would have a graph that trends in a general 45^o angle, but which would wiggle up and down from that straight line. As the canadian dollar goes through a period of losing value to the american dollar you would see the curve drop down, and as the canadian dollar goes through a period of gaining value to the american dollar (as it is currently?) then you would see the curve rise.My biggest concern here, though, is that you are getting caught in the forest of the calibration issue rather than looking at the tree of the correlation issue.If there was no correlation, there would be no way to calibrate it.Hope that helps.Enjoy.

---

we are limited in our ability to understand
by our ability to understand RebelAAmerican.Zen[Deist
{{{Buddha walks off laughing with joy}}}